1. Field of the Invention
The present invention relates to a medical image processing apparatus and method appropriate for comparison reading of medical images of an examined subject (patient), and a computer-readable recording having stored therein a program for causing a computer to execute the medical image processing method.
2. Description of the Related Art
In recent years, modalities using various techniques, such as an X-ray CT (Computed Tomography) apparatus, an ultrasonic (US) diagnosis apparatus, an MRI (Magnetic Resonance Imaging) apparatus, a PET (Positron Emission Tomography) apparatus and a SPET (Single-Photon Emission Tomography) apparatus, were used in medical image fields in addition to an X-ray radiography apparatus. As the speed of the modalities became higher, and the performance of the modalities improved to cope with multi-slice tomography or the like, it became possible to perform radiography on a plurality of regions of a patient in a series of operations. Accordingly, it is possible to obtain several hundred to several thousand highly-precise tomographic images of the patient. Further, a plurality of modalities came to be used to obtain medical images for a single candidate disease. Further, in diagnosis using images (image diagnosis), medical images obtained in past radiography are referred to in some cases. Therefore, comparison image reading became often performed. In comparison image reading, medical images obtained by a plurality of modalities and past medical images are compared with each other to make a diagnosis for one candidate disease. Hence, integrated diagnosis became possible.
In comparison image reading as described above, especially when the medical images are tomographic images, it is necessary to match the slice positions of the tomographic images obtained in different series of radiography, which were performed at different time from each other, and to display images of the same slice position. However, since several hundred to several thousand medical images are obtained in a series of radiography, the operation for matching the slice positions of slice images to be compared with each other imposes an extremely heavy burden on users.
Therefore, a method for automatically displaying images of the anatomically same slice position is proposed (please refer to Japanese Unexamined Patent Publication No. 2001-137230 (Patent Document 1)). In the method, a region of a specific organ, such as the lung field, heart and descending aorta, is extracted from a present medical image, which is a target of image reading, and a past medical image. Further, anatomical positions with respect to the body axis direction of a patient are matched between the present medical image and the past medical image based on an index with respect to the size and the morphology of the region extracted from the present medical image and the region extracted from the past medical image. Accordingly, images of the anatomically same slice position are automatically displayed. When the method disclosed in Patent Document 1 is used, an operation for matching the slice positions is not necessary. Therefore, it is possible reduce the burden on users in comparison image reading.
When image diagnosis is made in an actual field of medicine, diagnosis is often made by cooperation of a doctor in each clinical department, who requests image reading or diagnosis, (hereinafter, referred to as a requesting doctor) and a doctor or radiologist who is specialized in image reading (hereinafter, referred to as an image reading doctor). Specifically, after a radiographer obtains a medical image of a patient by performing radiography on the patient based on an examination order from the requesting doctor, the image reading doctor performs image reading on the medical image. The image reading doctor prepares an image reading report summarizing the result of image reading, and sends the image reading report to the requesting doctor. The requesting doctor reads a finding included in the image reading report. The requesting doctor makes a final diagnosis by referring to an image attached to the image reading report while taking various kinds of information, such as the result of questioning the patient and other examination results, into consideration.
As a conventional technique related to preparation of the image reading report and reference to the image reading report, a medical report preparation/reference apparatus is known (please refer to U.S. Patent Application Publication No. 20050226405 (Patent Document 2)). In the apparatus, when the image reading report is prepared, links are generated, based on an input operation by a user, in key character strings in the finding. The links are generated to clearly show correspondence between each of a plurality of images attached to the image reading report and a part of the content of the finding in the image reading report related to the respective images. The key character strings are linked to addresses of medical images that should be referred to together with the key character strings. Further, the key character strings are linked to image edit parameter, such as a window level, a window width and a magnification ratio. When the image reading report is referred to, if a person who views the image reading report clicks a character string in which a link is generated, a medical image linked to the character string is edited based on the image edit parameter, and displayed.
Further, a virtual endoscope technique is known. In the virtual endoscope technique, an image similar to an image obtained by using an endoscope is generated from a tomographic image obtained by radiography using a CT apparatus or the like. This technique is commonly used in North America especially in detection of tumors in large intestines, because tumors are detected only by CT radiography without examination using an endoscope. Further, a technique using the virtual endoscopic image has been proposed to perform safe and speedy surgery, examination or the like by using an endoscope.
For example, Japanese Unexamined Patent Publication No. 2006-198032 (Patent Document 3) discloses a technique of displaying a virtual endoscopic image of a peripheral region that does not appear in an endoscopic image when observation is performed by using an endoscope. The virtual endoscopic image of the peripheral region is superimposed on an endoscopic image obtained by the endoscope. The technique can assist a doctor in surgery using the endoscope. Further, Japanese Unexamined Patent Publication No. 2000-135215 (Patent Document 4) discloses a technique of generating a virtual endoscopic image of a duct, such as a bronchus. In this technique, the virtual endoscopic image is used to obtain, in advance, the movement path of an endoscope to a target point along the duct. A virtual endoscopic image along the path is used as a guide image when observation is performed actually by using an endoscope. The techniques disclosed in Patent Documents 3 and 4 can perform safe and speedy observation using an endoscope by displaying a virtual endoscopic image together with an endoscopic image.
However, in the method disclosed in Patent Document 1, the slice positions are matched based on the size of a region extracted from the present medical image, which is the target of image reading, and the size of a region extracted from the past medical image. Therefore, the accuracy of matching the positions is not sufficient. Consequently, it is necessary to manually correct the slice positions after the medical images are displayed. Further, in the method disclosed in Patent Document 2, a medical image correlated with the image reading report can be displayed by referring to the image reading report. However, in comparison image reading, a word that is correlated with a medical image to be compared with needs to be found in the image reading report. Therefore, the operation for searching the image reading report for the word is extremely troublesome. Further, when the medial image to be compared with is not correlated with any word in the image reading report, it is impossible to display the medical image.